Read­ing Time: 12 minutes

Adobe rolls out more gen­er­at­ive AI fea­tures to Illus­trat­or and Photoshop 

How to make Adobe Generative Fill and Expand less frustrating

Exper­i­ment­ing with selec­tions, con­text, and prompts can play a big role in get­ting a qual­ity res­ult. Make sure to keep in mind the size of the area you are gen­er­at­ing and con­sider work­ing in iter­at­ive steps, instead of try­ing to get the per­fect res­ult from a single prompt. Lead­ing enter­prises includ­ing the Coca-Cola Com­pany, Dick’s Sport­ing Goods, Major League Base­ball, and Mar­ri­ott Inter­na­tion­al cur­rently use Adobe Exper­i­ence Plat­form (AEP) to power their cus­tom­er exper­i­ence ini­ti­at­ives. Appar­ently, you can­’t use the new Gen­er­at­ive Fill fea­ture until you’ve shared some per­son­al identi­fy­ing inform­a­tion with the Adobe Behance cloud ser­vice. Behance users, by con­trast, will have already shared their con­fid­en­tial inform­a­tion with the ser­vice and be able to access the Pho­toshop Gen­er­at­ive Fill AI feature.

And with great power comes respons­ib­il­ity so Adobe says it wants to be a trus­ted part­ner for cre­at­ors in a way that is respect­ful and sup­port­ive of the cre­at­ive com­munity. Adobe Fire­fly gen­er­at­ive AI tools rid­ing shot­gun can unlock lim­it­less pos­sib­il­it­ies to boost pro­ductiv­ity and cre­ativ­ity. Every con­tent cre­at­or, solo­pren­eur, side hust­ler, and freel­ance artist has hit road­b­locks, maybe because of their skill level or per­haps a lack of time; it hap­pens. When build­ing a team isn’t pos­sible, Adobe Fire­fly gen­er­at­ive AI can help fill those gaps. Addi­tion­al cred­its can be pur­chased through the Cre­at­ive Cloud app, but only 100 more per month. That costs $4.99 a month if billed monthly or $49.99 if a full year is paid for up-front.

The recently launched GPU-accel­er­ated Enhance Speech, AI Audio Cat­egory Tag­ging and Filler Word Detec­tion fea­tures allow edit­ors to use AI to intel­li­gently cut and modi­fy video scenes. Instead, it main­tains that this update to its terms was inten­ded to cla­ri­fy its improve­ments to mod­er­a­tion pro­cesses. Due to the “explo­sion” of gen­er­at­ive AI, Adobe said it has had to add more human mod­er­a­tion to its con­tent sub­mis­sions review processes.

Will the stock be an AI winner?

Remove Back­ground is a good choice for those look­ing to build a com­pos­ite, as simply remov­ing the back­ground is all that is required. How­ever, for some Stock cus­tom­ers, they don’t want a back­ground; they require a dif­fer­ent one alto­geth­er. It brings new tools like the Gen­er­at­ive Shape Fill, so you can add detailed vec­tors to shapes using just a few descript­ive words. Anoth­er is a Text to Pat­tern fea­ture, whichen­ables the cre­ation of cus­tom­iz­able, scal­able vec­tor pat­terns. This update integ­rates AI in a way that sup­ports and amp­li­fies human cre­ativ­ity, rather than repla­cing it.

The part­ner­ship also aims to mod­ern­ize con­tent sup­ply chains using GenAI and Adobe Express to deploy innov­at­ive work­flows, allow­ing for a more diverse and col­lab­or­at­ive team to handle cre­at­ive tasks. While the com­pan­ies are yet to reveal fur­ther details about any products they will be releas­ing togeth­er, they did out­line the fol­low­ing four cross-com­pany integ­ra­tions that joint cus­tom­ers will be able to access. These work sim­il­arly to Adapt­ive Pre­sets, but they’ll pop up and dis­ap­pear depend­ing on what’s iden­ti­fied in your image. If a per­son is smil­ing, you’ll see Quick Actions relat­ing to whiten­ing teeth, mak­ing eyes pop, or real­ist­ic skin smooth­ing, for example. The new Adapt­ive Pre­sets use AI to scan your image and sug­gest pre­sets that suit the con­tent of the image best. While they can edit them to your lik­ing, they’ll adapt to what the AI thinks your image needs most.

Adobe Firefly

Illus­trat­or, Adobe’s vec­tor graph­ics edit­or, now includes Objects on Path, a fea­ture that allows users to quickly arrange objects along any path on their art­board. The soft­ware also boasts Enhanced Image Trace, which Adobe says improves the con­ver­sion of images to vec­tors. Adobe’s flag­ship image edit­ing soft­ware, Pho­toshop, received sev­er­al new features.

Around 90% of con­sumers report enhanced online shop­ping exper­i­ences thanks to AI. Key areas of improve­ment include product per­son­al­iz­a­tion, ser­vice recom­mend­a­tions, and the abil­ity to see vir­tu­al images of them­selves wear­ing products, with 91% stat­ing this would boost pur­chase con­fid­ence. Adobe made the announce­ment at the open­ing key­note of this year’s MAX con­fer­ence and plans to add this new Fire­fly gen­er­at­ive AI mod­el to Premiere Pro work­flows (more on those later).

By click­ing the but­ton, I accept the Terms of Use of the ser­vice and its Pri­vacy Policy, as well as con­sent to the pro­cessing of per­son­al data. Read our digit­al arts trends 2025 art­icle and our 3D art trends 2025 fea­ture for the latest tech, style and work­flow pre­dic­tions. “For best res­ults when using Gen Remove is to make sure you brush the object you’re try­ing to remove com­pletely includ­ing shad­ows and reflec­tion. Any leftover frag­ments, no mat­ter how small, will cause the AI to think it needs to attach a new object to that leftover piece. The GIP Digit­al Watch Obser­vat­ory team, con­sist­ing of over 30 digit­al policy experts from around the world, excels in the fields of research and ana­lys­is on digit­al policy issues. The team is backed by the cre­at­ive prowess of Cre­at­ive Lab Dip­lo and the tech­nic­al expert­ise of the Dip­lo tech team.

His­tor­ic­al invest­ment per­form­ances are no indic­a­tion or guar­an­tee of future suc­cess or per­form­ance. We make no rep­res­ent­a­tions or war­ranties regard­ing the advis­ab­il­ity of invest­ing in any par­tic­u­lar secur­it­ies or util­iz­ing any spe­cif­ic invest­ment strategies. Adobe has embed­ded AI tech­no­lo­gies into its exist­ing products like Pho­toshop, Illus­trat­or and Premiere Pro, giv­ing users more reas­ons to use its soft­ware, Durn said. Digit­al media and mar­ket­ing soft­ware firm Adobe (ADBE) impressed Wall Street ana­lysts with gen­er­at­ive AI innov­a­tions at the start of its Adobe Max con­fer­ence on Monday. You can now remove video back­grounds in Express, allow­ing you to apply the same edits to your con­tent wheth­er you’re using a photo or a video of a cut-out sub­ject. Adobe Express intro­duced a Dynam­ic Reflow Text tool, allow­ing you to eas­ily res­ize your Express artboards—using the latest gen­er­at­ive expand res­ize tool—and the text will dynam­ic­ally flow to fit the space you’ve created.

These include Dis­trac­tion Remov­al, which uses AI to elim­in­ate unwanted ele­ments from images, and Gen­er­at­ive Work­space, a tool for sim­ul­tan­eous ideation and concept devel­op­ment. The com­pany, which pro­duces soft­ware such as Pho­toshop and Illus­trat­or, unveiled over 100 new cap­ab­il­it­ies for its Cre­at­ive Cloud plat­form, many of which lever­age arti­fi­cial intel­li­gence to enhance con­tent cre­ation and edit­ing pro­cesses. Adobe, known for its cre­at­ive and mar­ket­ing tools, has announced a suite of new fea­tures and products at its annu­al MAX con­fer­ence in Miami Beach. Set to debut in beta form, the video expan­sion to the Fire­fly tool will integ­rate with Adobe’s flag­ship video edit­ing soft­ware, Premiere Pro. This integ­ra­tion aims to stream­line com­mon edit­or­i­al tasks and expand cre­at­ive pos­sib­il­it­ies for video professionals.

The com­pany’s latest Fire­fly Vec­tor AI mod­el is at the heart of these enhance­ments, prom­ising to sig­ni­fic­antly accel­er­ate cre­at­ive work­flows for graph­ic design­ers, fash­ion design­ers, interi­or design­ers or pro­fes­sion­al cre­at­ives. In a sep­ar­ate Adobe Com­munity post, a pro­fes­sion­al pho­to­graph­er says they use gen­er­at­ive fill “thou­sands of times per day” to “repair” their images. When Adobe deb­uted the Fire­fly-powered Gen­er­at­ive Remove tool in Adobe Light­room and Adobe Cam­era Raw in May as a beta fea­ture, it worked well much of the time. How­ever, Gen­er­at­ive Remove, now offi­cially out of its beta peri­od, has con­fus­ingly got­ten worse in some situ­ations. Adobe’s Gen­er­at­ive Fill and Expand tools can be frus­trat­ing, but with the right tech­niques, they can also be very useful.

That’s a key dis­tinc­tion, as Photoshop’s exist­ing AI-based remov­al tools require the edit­or to use a brush or selec­tion tool to high­light the part of the image to remove. In pre­views, Adobe demon­strated how the tool could be used to remove power lines and people from the back­ground without mask­ing. The third AI-based tool for video that the com­pany announced at the start of Adobe Max is the abil­ity to cre­ate a video from a text prompt. While text to video is Adobe’s video vari­ation of cre­at­ing some­thing from noth­ing, the com­pany also noted that it can be used to cre­ate over­lays, anim­a­tions, text graph­ics or B‑roll to add to exist­ing cre­ated-with-a-cam­era video. It’s based on Gen­er­at­ive Fill, but rather than repla­cing a user-selec­ted por­tion of an image with AI-gen­er­ated con­tent, it auto­mat­ic­ally detects and replaces the back­ground of the image.

Behind the scenes: How Paramount+ used Adobe Firefly generative AI in a social media campaign for the movie IF – the Adobe Blog

Behind the scenes: How Para­mount+ used Adobe Fire­fly gen­er­at­ive AI in a social media cam­paign for the movie IF.

Pos­ted: Mon, 09 Dec 2024 08:00:00 GMT [source]

The Gen­er­at­ive Shape Fill tool is powered by the latest beta ver­sion of Fire­fly Vec­tor Mod­el which offers extra speed, power and pre­ci­sion. It includes text-to-image and gen­er­at­ive fill, video tem­plates, stock music, image and design assets, and quick-action edit­ing tools to help you cre­ate con­tent eas­ily on the go. Once you have cre­ated con­tent, you can plan, pre­view, and pub­lish it to Tik­Tok, Ins­tagram, Face­book, and Pin­terest without leav­ing the app. Recog­nising the grow­ing need for effi­cient col­lab­or­a­tion in cre­at­ive work­flows, Adobe announced the gen­er­al avail­ab­il­ity of a new ver­sion of Frame.io.

Some of you might leave since you can’t pay the annu­al fee upfront or afford the monthly increase. We can hardly be bothered as we need more cash to come up with more and more AI-related gim­micks that pho­to­graph­ers like you will hardly ever use. It’s not so much that Adobe’s tools don’t work well, it’s more the man­ner of how they’re not work­ing well — if we weren’t try­ing to get work done, some of these res­ults would be really funny. In the case of the Bit­coin thing, it just seems like it’s try­ing to replace the painted pixels with some­thing sim­il­ar in shape to the detec­ted “object” the user is try­ing to remove. Last week, I struggled to get any of Adobe’s gen­er­at­ive or con­tent-aware tools to extend a back­ground and cov­er an area for a thumb­nail I was work­ing on for our You­Tube chan­nel. Pre­vi­ous to the updates last year, the tasks I asked Pho­toshop to handle were done quickly and without issue.

Adobe is listen­ing to feed­back and mak­ing tweaks, but AI incon­sist­en­cies point toward a broad­er issue. Gen­er­at­ive AI is still a nas­cent tech­no­logy and, clearly, not one that exclus­ively improves with time. Some­times it gets worse, and for those with an AI-reli­ant work­flow, that’s a prob­lem that under­cuts the util­ity of gen­er­at­ive AI tools altogether.

Adobe’s new AI tool can edit 10,000 images in one click

The Adobe Fire­fly Video Mod­el — now avail­able in lim­ited beta at Firefly.Adobe.com — brings gen­er­at­ive AI to video, mark­ing the next advance­ment in video edit­ing. It allows users to cre­ate and edit video clips using simple text prompts or images, help­ing fill in con­tent gaps without hav­ing to reshoot, extend or reframe takes. It can also be used to cre­ate video clip pro­to­types as inspir­a­tion for future shots. Adobe unveiled its Fire­fly Video Mod­el last month, pre­view­ing a vari­ety of new gen­er­at­ive AI video fea­tures. Today, the Fire­fly Video Mod­el has offi­cially launched in pub­lic beta and is the first pub­licly avail­able gen­er­at­ive video mod­el designed to be com­mer­cially safe.

That cov­ers the main set of con­trols which over­lay the right of your image – but there is a smal­ler set of con­trols on the left that we must explore as well. Back up to the set of three con­trols, the middle option allows you to ini­ti­ate a Down­load of the selec­ted image. As Fire­fly begins pre­par­ing the image for down­load, a small over­lay dia­log appears.

There are also Text to Pat­tern, Style Ref­er­ence and more work­flow enhance­ments that can ser­i­ously speed up tedi­ous design and draw­ing tasks enabling design­ers to dive deep­er into their work. Everything from the ini­tial con­cep­tion of an idea through to final pro­duc­tion is get­ting a help­ing hand from AI. If you do hap­pen to have a team around you, fea­tures like brand kits, co-edit­ing, and com­ment­ing will aid in faster, more seam­less collaboration.

Adobe is using AI to make the cre­at­ive pro­cess of design­ing graph­ics much easi­er and quick­er, … [+] leav­ing users of pro­grams like Illus­trat­or and Pho­toshop free to spend more time with the cre­at­ive pro­cess. Adobe has some lan­guage included that appears to be a hol­d­over from the ini­tial launch of Fire­fly. For example, the com­pany stip­u­lates that the Cred­it con­sump­tion rates above are for what it calls “stand­ard images” that have a res­ol­u­tion of up to 2,000 by 2,000 pixels — the ori­gin­al max­im­um res­ol­u­tion of Fire­fly gen­er­at­ive AI. Along that same line of think­ing, Adobe says that it hasn’t provided any notice about these changes to most users since it’s not enfor­cing its lim­its for most plans yet.

To date, Fire­fly has been used by numer­ous Adobe enter­prise cus­tom­ers to optim­ize work­flows and scale con­tent cre­ation, includ­ing PepsiCo/Gatorade, IBM, Mat­tel, and more. This con­cern stems from the idea that even­tu­ally, AI-gen­er­ated con­tent will make up a large por­tion of train­ing data, and the res­ults will be AI slop – wonky, erro­neous or unus­able images. The self-per­petu­at­ing cycle would even­tu­ally render the tools use­less, and the qual­ity of the res­ults would be degraded. It’s espe­cially wor­ri­some for artists who feel their unique styles are already being co-opted by gen­er­at­ors, res­ult­ing in ongo­ing law­suits over copy­right infringe­ment concerns.

• The samples shared in the announce­ment show a pretty power­ful mod­el, cap­able of under­stand­ing the con­text and provid­ing coher­ent generations.
• IBM is exper­i­ment­ing with Adobe Fire­fly to optim­ize work­flows across its mar­ket­ing and con­sult­ing teams, focus­ing on devel­op­ing reli­able AI-powered cre­at­ive and design outputs.
• Adobe has also improved its exist­ing Fire­fly Image 3 Mod­el, claim­ing it can now gen­er­ate images four times faster than pre­vi­ous versions.
• It also emerged that Can­on, Nikon and Leica will sup­port its Cam­era to Cloud (C2C) fea­ture, which allows for dir­ect uploads of pho­tos and videos to Frame.io.

But as the Len­ovo example shows, there’s a lot of care­ful ground­work required to safely har­ness the poten­tial of this new tech­no­logy. If you look at the amount of con­tent that we need to achieve end-to-end per­son­al­iz­a­tion, it’s pretty astro­nom­ic­al. To give you an example, we just launched a cam­paign for four products across eight mar­ket­ing chan­nels, four lan­guages, and three vari­ations. Speed­ing up con­tent deliv­ery in this way means that teams are then able to adjust and fine-tune the exper­i­ence in real-time as trends or needs change.

How­ever, at the moment, these latest gen­er­at­ive AI tools, many of which were speed­ing up their work­flows in recent months, are now slow­ing them down thanks to strange, mis­matched, and some­times baff­ling res­ults. “The gen­er­at­ive fill was almost per­fect in the pre­vi­ous ver­sion of Pho­toshop to com­plete this task. Since I updated to the new­est ver­sion (26.0.0), I get very absurd res­ults,” the user explains. Since the update, gen­er­at­ive fill adds objects to a per­son, includ­ing a rab­bit and let­ters on a person’s face. Illus­trat­or and Pho­toshop have received GenAI tools with the goal of improv­ing user exper­i­ence and allow­ing more free­dom for users to express their cre­ativ­ity and skills. Our com­mit­ment to evolving our assess­ment approach as tech­no­logy advances is what helps Adobe bal­ance innov­a­tion with eth­ic­al responsibility.

We gath­er data from the best avail­able sources, includ­ing vendor and retail­er list­ings as well as oth­er rel­ev­ant and inde­pend­ent reviews sites. And we pore over cus­tom­er reviews to find out what mat­ters to real people who already own and use the products and ser­vices we’re assess­ing. Ghost­G­PT can also be used for cod­ing, with the blog post not­ing mar­ket­ing related to mal­ware cre­ation and exploit devel­op­ment. Mal­ware authors are increas­ingly lever­aging AI cod­ing assist­ance, and tools like Ghost­G­PT, which lack the typ­ic­al guard­rails of oth­er large lan­guage mod­els (LLMs), can save crim­in­als time spent jail­break­ing main­stream tools like Chat­G­PT. Media Intel­li­gence auto­mat­ic­ally recog­nises clip con­tent, includ­ing people, objects, loc­a­tions, cam­era angles, cam­era type and more. This allows edit­ors to simply type out the clip type needed in the new Search Pan­el, which dis­plays inter­act­ive visu­al res­ults, tran­scripts, and oth­er metadata res­ults from across an entire project.

An Adobe rep­res­ent­at­ive says that today, it does have in-app noti­fic­a­tions in Adobe Express — an app where cred­its are enforced. Once Adobe does enforce Gen­er­at­ive Cred­its in Pho­toshop and Light­room, the com­pany says users can abso­lutely expect an in-app noti­fic­a­tion to that effect. As part of the ori­gin­al story below, Peta­Pixel also added a line stat­ing that in-app noti­fic­a­tions are being used in Adobe Express to let users know about Gen­er­at­ive Cred­its use. Look­ing ahead, Adobe fore­cast fisc­al fourth-quarter rev­en­ue of between $5.5 bil­lion and $5.55 bil­lion, rep­res­ent­ing growth of between 9% to 10%.

In addi­tion, Adobe is adding a neat fea­ture to the Remove tool, which lets you delete people and objects from an image with ease, like Google’s Magic Eraser. With Dis­trac­tion Remov­al, you can remove cer­tain com­mon ele­ments with a single click. For instance, it can scrub unwanted wires and cables, and remove tour­ists from your travel pho­tos. Adobe is join­ing sev­er­al oth­er play­ers in the gen­er­at­ive AI (GAI) space by rolling out its own mod­el. The Fire­fly Video Mod­el is power­ing a num­ber of fea­tures across the com­pany’s wide array of apps.

It works great for remov­ing cables and wires that dis­tract from a beau­ti­ful sky­scape. This really begins with defin­ing our brand and chan­nel guidelines as well as per­so­nas in order to gen­er­ate con­tent that is on-brand and sup­ports per­son­al­iz­a­tion across our many seg­ments. The rap­id adop­tion of gen­er­at­ive AI has cer­tainly cre­ated chaos inside and out­side of the cre­at­ive industry. Adobe has tried to mit­ig­ate some of the con­fu­sion and con­cerns that come with gen AI, but it clearly believes this is the way of the future. Even though Adobe cre­at­ors are excited about spe­cif­ic AI tools, they still have ser­i­ous con­cerns about AI’s over­all impact on the industry.

One cap­ab­il­ity gen­er­ates visu­al assets sim­il­ar to the one high­lighted by a design­er. The oth­ers can embed new objects into an image, modi­fy the back­ground and per­form related tasks. Some of the cap­ab­il­it­ies are rolling out to the company’s video edit­ing applic­a­tions. The oth­ers will mostly become avail­able in Adobe’s suite of image edit­ing tools, includ­ing Pho­toshop. For pho­to­graph­ers not opposed to gen­er­at­ive AI in their photo edit­ing work­flows, Gen­er­at­ive Remove and oth­er gen­er­at­ive AI tools like Gen­er­at­ive Fill and Gen­er­at­ive Expand have become indispensable.

 Email This Post